基于深度学习进化的超大规模业务请求与服务功能链高效匹配

Efficient Matching of Ultra-Large-Scale Business Requests to Service Function Chains Based on Deep Learning Evolution

摘要: 随着5G技术与电子商务平台的深度融合，实现业务请求与服务功能链(SFC)之间的高效匹配，对于保障电子商务各类交易的平稳运行与顺利实施具有深远的研究意义。如何为这些多样化的业务请求匹配到满足其特定需求的SFC，已成为当前电子商务领域亟待解决的重要问题。为应对上述挑战，本文深入剖析了电子商务业务请求与NFC的固有特征，并据此构建了两者之间的精准匹配模型。本文创新性地设计了一个基于深度学习进化的优化算法，旨在高效求解所建立的匹配模型。该算法通过巧妙引入卷积神经网络处理业务请求特征并求解适应度函数为遗传算法提供优化的搜索环境，显著提升了匹配模型的精度与效率，并有效降低了计算复杂度，从而为电子商务环境中的服务功能链匹配问题提供了一种新颖且实用的解决方案。

Abstract: The deep integration of 5G technology with e-commerce platforms underscores the profound research significance of achieving efficient matching between business requests and service function chains (SFC) to ensure the smooth operation and successful implementation of diverse e-commerce transactions. A pivotal challenge in the current e-commerce domain is matching these diversified business requests to SFC that meet their specific needs. To address this challenge, this paper conducts an in-depth analysis of the inherent characteristics of e-commerce business requests and SFCs, and subsequently constructs a precise matching model between them. Innovatively, we design an optimization algorithm based on deep learning evolution, tailored to efficiently solve the established matching model. By ingeniously incorporating convolutional neural networks (CNNs) to process business request features and solving the fitness function to provide an optimized search environment for the genetic algorithm, the proposed algorithm significantly enhances the accuracy and efficiency of the matching model while effectively reducing computational complexity. This paper thus presents a novel and practical solution to the problem of service function chain matching in e-commerce environments.